In Brief
At DePIN Day Dubai, the panel "Can DePIN Make AI More Scalable, Secure and Trustless?" gathered the pioneers of decentralized infrastructure to tackle a hot question: Can we build AI that’s decentralized and powerful enough to compete with centralized giants? From GPUs and bandwidth to verifiability and trust, the panelists explored how DePIN is shaping the future of AI at scale — and what challenges remain before it can match (or even beat) traditional cloud.
🗣️ Evgeny Ponomarev: Building the Cloudless Alternative
Evgeny Ponomarev of Fluence framed his vision of a "cloudless" internet. Fluence aggregates compute from professional providers worldwide, offering decentralized cloud services that rival AWS and Google. Evgeny emphasized that decentralized AI isn’t just about replacing servers — it’s about rethinking the economics of compute. By crowdsourcing hardware and rewarding contributors with tokens, Fluence aims to break the stranglehold of hyperscalers and deliver affordable, open infrastructure for Web3-native AI.
🗣️ Kate Lane: Aggregating GPU Power for AI Builders
Kate Lane from io.net put the spotlight on decentralized GPU networks. Io.net is building a global pool of GPUs and CPUs by aggregating resources from data centers and individual providers. For AI builders facing skyrocketing cloud costs or limited access to high-performance GPUs, io.net is positioning itself as a decentralized alternative to AWS. With pay-as-you-go models and lower overheads, Kate argued, DePIN offers a path to cheaper, more accessible compute — a game-changer for GenAI projects that can’t afford to be locked into Big Cloud contracts.
🗣️ Ayush Ranjan: Cutting Costs and Latency for Real-Time AI
Ayush Ranjan of Huddle01 made a compelling case for decentralized bandwidth networks. Huddle01 is tackling the notorious “egress cost” problem: hyperscalers often charge up to 80x markup on data-out fees, which crushes AI apps with real-time voice or video. Ayush explained how decentralizing bandwidth — and moving servers closer to end users — can reduce these costs by up to 90% while slashing latency. This, he argued, is a critical advantage for AI-powered voice assistants and real-time apps that can’t afford to wait seconds for a response.
🗣️ Prabal Banerjee: Enabling Verifiable, Modular AI Infrastructure
Prabal Banerjee from Avail zoomed out to the blockchain backbone that supports DePIN. Avail is building horizontally scalable blockchain infra, essential for making decentralized AI workloads verifiable and modular. Prabal highlighted that today’s AI models — built on reputation and closed systems — require social capital and trust. Verifiability, he argued, is the key to unlocking smaller players, enabling AI at the edge and scaling trustless systems that users can depend on.
The Challenge
While excitement around DePIN-powered AI is palpable, panelists agreed on several challenges:
Hardware Fragmentation: High-end GPUs (like Nvidia’s H100) are still scarce at the edge, complicating distributed AI training.
Service-Level Agreements: Decentralized networks struggle to offer the 99% uptime that enterprise AI buyers demand.
Data Transfer Costs: Hyperscalers’ egress fees remain a major barrier to scaling decentralized AI cost-effectively.
Verifiability: Ensuring outputs are reliable and tamper-proof is still an emerging frontier for DePIN-based AI.
The Solution
To bridge these gaps, the panelists outlined several strategies:
Token Incentives: Aligning contributors through staking, block rewards and revenue-sharing to build robust, global networks.
Edge Deployment: Bringing models closer to users to cut egress costs and boost responsiveness — especially in underserved regions.
SLA Standardization: Developing DePIN-native mechanisms to meet enterprise-grade uptime and performance.
Verifiable Infrastructure: Integrating ZK proofs and other cryptographic tools to guarantee trustless, auditable outputs.
Going Beyond
The panelists agreed: DePIN is more than an alternative to Big Cloud — it’s a chance to rethink how AI is built, priced and delivered. From self-replicating AI agents to decentralized gaming and satellite networks in low-Earth orbit, the future is about unlocking user ownership, transparency and global participation. As token models mature and hardware supply chains catch up, decentralized AI could soon rival — and maybe surpass — its centralized rivals.
🌐 Join the DePIN Day Movement
Be part of the next chapter in decentralized infrastructure! Whether you’re a visionary founder, a curious developer, a bold investor, or a pioneering researcher — DePIN Day is your stage. Explore, connect, and build the future with us.
👉 Apply now to become a speaker, sponsor or participant at our upcoming events
